summary.cv.grpnet: Summarize cv.grpnet Fits

In grpnet: Group Elastic Net Regularized GLMs and GAMs

Summarize cv.grpnet Fits

Description

Obtain fitted values, variable activations, variable importances from a fit group elastic net (grpnet) object.

Usage

## S3 method for class 'cv.grpnet'
summary(object, ...)

Arguments

object	Object of class "cv.grpnet"
...	Additional arguments passed to predict.cv.grpnet function.

Details

Passes the training data to the predict.cv.grpnet function four times: two predictions types (fit and imp) by two lambda values (lambda.1se and lambda.min).

Value

family	Name of exponential family/loss function.
penalty	Name of L1 group penalty.
nobs	Number of observations.
ngroups	Number of coefficient groups.
lambda	Vector of lambda values (1se and min).
dev.ratio	Vector of proportions of null deviance explained (1se and min).
fit	Fitted values obtained from predict.cv.grpnet.
act	Variable activations obtained from predict.cv.grpnet.
imp	Variable importances obtained from predict.cv.grpnet.

Note

The variables activations are defined as: act <- abs(imp) > 0.0

For most response families, the returned components fit, act, and imp are matrices of dimension K by 2.

For families with multivariate responses (i.e., multigaussian and multinomial), these components are arrays of dimension K by L by 2 where L is the dimension of the response.

Author(s)

Nathaniel E. Helwig <[e-mail]>

References

Helwig, N. E. (2025). Versatile descent algorithms for group regularization and variable selection in generalized linear models. Journal of Computational and Graphical Statistics, 34(1), 239-252. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1080/10618600.2024.2362232")}

See Also

cv.grpnet for k-fold cross-validation of lambda

Examples

# see 'cv.grpnet' for summary examples
?cv.grpnet

grpnet documentation built on May 2, 2026, 9:07 a.m.